Video
Introducing Observe by Snowflake
Jeremy Burton, Snowflake GM of Observability, introduces a new model for observability built natively on the Snowflake lakehouse.
Observeby Snowflake
AI-powered observability at scale. Unify logs, metrics and traces on a cost-efficient Snowflake lakehouse, using an integrated AI SRE that helps resolve incidents faster.
BENEFITS
Troubleshoot fasterat lower cost
Programmable AI SRE
Run AI-driven observability where you work
The Observe AI SRE, CLI and MCP Server support developers, engineers, DevOps and SREs where they are, enabling custom agentic workflows.
Context Engineered for Accuracy
Get faster, more accurate answers grounded in your data
The Observability Context Graph maps semantics and relationships across logs, metrics and traces, extending to code and business context — so you get faster, more accurate reasoning.
Cost-Efficient Scale
Cut observability costs while scaling and retaining more data
Observe’s Telemetry Lakehouse Foundation, built on Snowflake, offers low-cost cloud storage and compute-storage separation to ingest, store and analyze more data at lower cost.
use cases
Observability engineered for scale
INVESTIGATE AND ROOT CAUSE FASTER
Reduce MTTR with AI SRE
- Investigate incidents, find root causes and get suggested fixes using natural language.
- Get more accurate answers with AI SRE, which applies context from the Observability Context Graph.
- Fewer engineers need to respond to incidents, and when they do, there’s no need to jump between tools and dashboards.
- Troubleshoot using AI SRE in the Observe UI or run agentic workflows via CLI or MCP Server with Claude or Cursor.
Search logs at scale
Analyze log data at a fraction of the cost
- Search and analyze logs without worrying about indexing, data tiering or retention limits.
- Ingest log data in any format — structured, semi-structured or unstructured.
- Generate unlimited log-derived metrics at no extra cost.
END-TO-END SERVICE VISIBILITY
Go from service issues to root cause in seconds
- Native OpenTelemetry provides automatic instrumentation and unified service discovery.
- Capture every trace without sampling or data loss for full-fidelity investigation.
- Move seamlessly from traces to correlated logs, metrics and infrastructure data to reduce manual effort.
REAL-TIME SYSTEM HEALTH
Correlate infrastructure signals to understand system behavior
- Capture metrics across cloud, Kubernetes, containers and serverless from 400+ pre-built integrations.
- Skip dashboard-building and reduce alert noise with out-of-the-box visualizations and contextual alerts.
- Pivot easily between infrastructure dashboards and correlated logs to troubleshoot and resolve incidents.
OPTIMIZE AI WORKLOADS
Operate AI applications and agents with confidence
- Trace AI requests from prompt to response across multi-step agentic workflows.
- Analyze token consumption to spot usage patterns and see where you can trim spend.
- Connect AI application performance to underlying infrastructure and services.
Resources
Explore more inObservability
GET STARTED
Take the next step with Observe by Snowflake
Start your free trial
Performance at scale
Faster troubleshooting
Lower cost with an open data lake
Snowflake for Observability
Frequently Asked Questions
Your top questions about getting started with Observe by Snowflake, answered here.
How does Observe relate to Snowflake?
Observe by Snowflake is an AI-powered observability platform built natively on a Snowflake lakehouse. It extends Snowflake’s core capabilities — compute-storage separation, cloud object storage and columnar analytics, with optimizations for real-time telemetry ingestion and analysis. Observe lets customers apply Snowflake credits toward Observe usage and manage observability data alongside the rest of their enterprise data.
Do I need to be a customer of Snowflake’s AI Data Cloud to use Observe?
No. Observe is available as a standalone solution, and you do not need to be an existing Snowflake customer to use it. While Observe is built on Snowflake’s platform, no separate Snowflake purchase is required.
Who is Observe built for?
Observe serves the teams responsible for building and running software applications: SREs and DevOps engineers troubleshooting incidents, application developers debugging performance and errors, and product and support teams who need visibility without writing extensive queries. The Observe UI, AI SRE, MCP server, and CLI meet each of these users in the tools they already work in.
As teams adopt agentic workflows, Observe is built to serve those agents as well. Observe provides MCP, CLI and API building blocks that allow agentic workflows to query observability data directly, so when a human or an agent investigates an incident, they have the relevant telemetry and context to act on.
How is Observe different from other observability tools?
There are two key architectural differences. First, Observe stores telemetry in a single Telemetry Lakehouse Foundation that separates compute from storage, lowering cost at scale and removing the retention limits common in index-based tooling. Second, the Observability Context Graph automatically connects related signals across logs, metrics, traces, code, and business context, so AI agents and engineers investigate from one unified model instead of pivoting between disconnected dashboards.
How does Observe help lower observability costs?
Observe’s Telemetry Lakehouse Foundation, built on Snowflake, helps customers cost-effectively ingest, retain and analyze higher volumes of telemetry data. Because data lands in low-cost cloud storage and stays hot, there’s no archiving, storage tiering or rehydrating cold data. Consolidating logs, metrics and traces at scale can often eliminate the need for multiple observability tools, reducing licensing costs and operational overhead.
What is an AI SRE?
An AI SRE is an AI-powered assistant that helps engineering teams troubleshoot faster by correlating telemetry data, surfacing root causes, and suggesting next steps during incidents. Observe’s AI SRE builds on this by grounding its analysis in your actual environment through the Observability Context Graph, which maps the relationships across your services, infrastructure, logs, metrics, and traces. You can access the AI SRE through a chat interface in Observe or directly from your coding agent, such as Claude or Cursor, via the MCP server.
How does Observe work with open standards like OpenTelemetry and Apache Iceberg™?
Observe natively ingests OpenTelemetry (OTel) data, so teams can use existing OTel instrumentation and collectors, including the Observe Agent, without relying on proprietary agents. For storage, Observe supports Apache Iceberg, storing observability data in open table formats that remain accessible outside the platform. This gives organizations flexibility to analyze telemetry through Observe or with any Iceberg-compatible engine, helping reduce vendor lock-in while preserving access to their data.
